Complex functions form the backbone of everything you'll encounter in complex analysis. When you're asked to evaluate contour integrals, find residues, or analyze singularities, you need to recognize which type of function you're dealing with and what properties it carries. The exponential, logarithmic, trigonometric, and rational functions aren't just formulas to memorize—they're the building blocks that determine whether a function is entire, meromorphic, or multi-valued, and that distinction drives nearly every problem you'll solve.
You're being tested on your ability to connect function definitions to their analytic properties: holomorphicity, periodicity, branch behavior, and singularity structure. Don't just memorize that ez is entire or that log(z) needs a branch cut—understand why these properties emerge and how they affect integration paths, series expansions, and transformations. Master the underlying mechanisms, and you'll handle any variation the exam throws at you.
The Exponential Family: Functions Built from ez
The exponential function is the engine driving most of complex analysis. Because ez is entire and periodic in the imaginary direction, it generates the trigonometric and hyperbolic functions while creating the multi-valuedness that makes logarithms tricky.
Exponential Function
Defined as ez=ex+iy=ex(cosy+isiny)—the only function that equals its own derivative in the complex plane
Periodic with period 2πi, meaning ez+2πi=ez; this periodicity is the root cause of logarithmic branch cuts
Entire function (holomorphic everywhere)—no singularities, no poles, making it the cleanest function to work with in contour integration
Trigonometric Functions
Defined via exponentials:sin(z)=2ieiz−e−iz and cos(z)=2eiz+e−iz—these definitions extend real trig functions to the complex plane
Entire functions with period 2π, but unlike their real counterparts, they are unbounded; ∣sin(z)∣ can grow arbitrarily large
Essential for Fourier analysis and signal processing; their zeros occur exactly at integer multiples of π
Hyperbolic Functions
Defined as sinh(z)=2ez−e−z and cosh(z)=2ez+e−z—note the relationship: sin(iz)=isinh(z)
Entire functions with purely imaginary period 2πi, connecting them directly to trigonometric functions through rotation
Critical in physics applications, particularly hyperbolic geometry, relativity, and solutions to the wave equation
Compare: Trigonometric vs. Hyperbolic functions—both are entire, both derive from ez, but trig functions use eiz (rotation) while hyperbolic functions use real exponentials (scaling). If an exam asks you to relate sin(iz) to hyperbolic functions, use sin(iz)=isinh(z).
Multi-Valued Functions: When One Answer Isn't Enough
Some complex functions naturally produce multiple outputs for a single input. The logarithm and fractional powers inherit multi-valuedness from the argument function, requiring branch cuts to make them single-valued and analytic.
Logarithmic Function
Defined as log(z)=ln∣z∣+iarg(z)—the argument arg(z) is only determined up to multiples of 2π, creating infinitely many values
Requires a branch cut (typically along the negative real axis) to restrict to a single-valued function; the principal branch uses −π<arg(z)≤π
Inverse of the exponential, essential for solving ew=z and for defining complex powers via za=ealog(z)
Branch Cuts and Riemann Surfaces
Branch cuts are artificial barriers in the complex plane that prevent paths from circling around branch points and switching between function values
Riemann surfaces provide a geometric resolution—a multi-sheeted surface where the function becomes single-valued by "unfolding" the plane
Essential for understanding topology of functions like z, log(z), and z1/n; the number of sheets equals the number of function values
Compare:log(z) vs. z—both are multi-valued, but log(z) has infinitely many branches (one for each 2πk) while z has exactly two. Riemann surface for log(z) has infinitely many sheets; for z, just two.
Holomorphic Hierarchy: Entire vs. Meromorphic
Understanding where a function is analytic determines which theorems apply. Entire functions are holomorphic everywhere; meromorphic functions allow isolated poles but nothing worse.
Entire Functions
Holomorphic on all of C—examples include polynomials, ez, sin(z), and cos(z)
Power series representationf(z)=∑n=0∞anzn converges for all z; the radius of convergence is infinite
Liouville's theorem applies: a bounded entire function must be constant—this proves the Fundamental Theorem of Algebra
Meromorphic Functions
Holomorphic except at isolated poles—points where the function blows up like (z−z0)−n for some positive integer n
Expressible as ratios of entire functions: if f(z)=g(z)/h(z) with g,h entire, then f is meromorphic with poles at zeros of h
Central to residue calculus; the residue theorem requires identifying poles and computing residues for contour integration
Rational Functions
Defined as f(z)=Q(z)P(z) where P and Q are polynomials—the simplest meromorphic functions
Poles occur at zeros of Q(z), with order equal to the multiplicity of the zero; partial fractions decomposition reveals the pole structure
Fundamental for contour integration—most residue calculations involve rational functions or functions reducible to them
Compare: Entire vs. Meromorphic—entire functions have no singularities at all; meromorphic functions allow poles but not essential singularities. ez is entire; 1/z is meromorphic; e1/z is neither (essential singularity at z=0).
Building Blocks: Power Functions and Transformations
These fundamental functions serve as the raw materials for series expansions and geometric mappings. Power functions generate Taylor and Laurent series; Möbius transformations preserve the analytic structure of the extended plane.
Power Functions
Defined as f(z)=zn for integer n—for positive n, entire; for negative n, meromorphic with a pole at the origin
Building blocks for series: Taylor series use non-negative powers; Laurent series include negative powers to capture pole behavior
Conformal except at critical points—zn multiplies angles by n, so it's not conformal at z=0 unless n=1
Möbius Transformations
Defined as f(z)=cz+daz+b with ad−bc=0—also called linear fractional transformations
Bijective and angle-preserving (conformal), mapping circles and lines to circles and lines; they form a group under composition
Act on the extended complex planeC∪{∞}, with f(−d/c)=∞ and f(∞)=a/c; essential for understanding the Riemann sphere
Compare: Power functions vs. Möbius transformations—zn is entire (for n≥0) but not injective; Möbius transformations are always bijective but have exactly one pole. Use power functions for local behavior; use Möbius transformations for global mappings.
Quick Reference Table
Concept
Best Examples
Entire functions
ez, sin(z), cos(z), polynomials
Meromorphic functions
Rational functions, tan(z), cot(z)
Multi-valued functions
log(z), z, z1/n
Require branch cuts
log(z), fractional powers, inverse trig
Self-derivative property
ez (unique among non-trivial functions)
Periodic in imaginary direction
ez (period 2πi)
Conformal mappings
Möbius transformations, ez, power functions (away from critical points)
Series building blocks
Power functions zn
Self-Check Questions
Which functions from this guide are entire, and what property do they all share regarding their power series representation?
Compare sin(z) and sinh(z): how are their definitions related, and what identity connects them?
Why does log(z) require a branch cut while ez does not? What property of ez causes this asymmetry?
If you're given a rational function f(z)=z3−zz2+1, how would you identify its poles and determine whether it's meromorphic or has worse singularities?
Explain how Möbius transformations differ from power functions in terms of injectivity and behavior at infinity—when would you use each type of mapping?